Proof of Stake or Work
The two main consensus mechanisms
The heart of a digital asset network is a consensus mechanism that makes the participants of the network agree on a shared state of the blockchain ledger. It regulates new units of supply as well as the different transactions that are made on the blockchain. Many different blockchains exist, with many more to come, with differing consensus mechanisms.
Understanding the differences can help give insight into different crypto assets and can help in evaluating which Layer-1 Blockchains are interesting to invest and participate in.The two most widely used consensus mechanisms are Proof of Work (PoW) and Proof of Stake (PoS), and they both regulate the process in which transactions between users are verified and added to a blockchain’s public ledger, all without a central party’s help.
Proof of Work
Proof of Work, coined in a scientific paper of Markus Jakobsson and Ari Juels in 1999, describes the process of having to perform a certain amount of work before some type of transaction gets concluded. Early use of this concept was a means to mitigate email spam where a user can perform a small task but the computing power and/or resources for mass emailing would be less viable. Ultimately, Satoshi Nakamoto, applied Proof of Work technology in Bitcoin in 2009.
PoW algorithms such as Bitcoin, determine who can adjust the ledger through a race of computation, where participants (miners) expend energy in order to have a chance to propose a new valid block with transactions to the blockchain. In order to create a new block, miners on a PoW network compete against each other to solve complex mathematical problems in a process called hashing. The trick with these mathematical problems is that they are very hard to solve, but easy to verify, ideal properties to secure the network with computational power.
Proof of Stake
Proof of Stake (PoS) is a modification of PoW that was introduced in 2012 for blockchains. Rather than relying on computationally expensive calculations to nominate a block producer, PoS decides the block producer based on the ownership of the coin of the network. A set of factors determined by the protocol elects a node, which job is then to verify the validity of the transactions in the block, sign it and propose it for validation in the network.
Proponents of PoS often argue that executing consensus this way is more efficient and uses a lot less energy to reach consensus with little trade-offs. While PoW mining requires computationally expensive problems to be solved, PoS systems use energy as well by needing to use specialized hardware (GPUs) and other computing devices in combination with active internet connections to produce blocks. Furthermore, because of the competitive nature of blockchain block producers, both systems will use the cheapest form of electricity, often waste electricity or renewables (pdf).
Comparing the consensus mechanisms
While evaluating differing consensus mechanisms, a key area of comparison is the trade-offs in regards to the blockchain trilemma: Security, Decentralization and Scalability. These properties are not the only valuable attributes of a blockchain, however you can argue this is a good starting point. Other areas such as for example, governance mechanisms, interoperability, data availability, coin distribution/Gini coefficients, developer activity, user activity, lock-in effects, first-mover advantages, Nakamoto coefficient, energy usage, amongst other attributes, are important as well to get a finer outlook of different blockchains with similar consensus mechanisms. For now, we will focus on the trilemma.
Scalability
Scalability is the term used in the blockchain context to refer to the ability to be used in a range of capabilities in a high number of occurrences, critical for blockchain adoption to happen. Increased adoption will only increase the demand of the blockspace and thus heighten its importance. Simply increasing the capacity of a ‘block’ of the blockchain to be able to carry more transactions is not a viable solution. Bigger blocks mean exponentially bigger blockchains over time, which results in less validators and ultimately less decentralization, underscoring the blockchain trilemma.
Off-chain scaling is the main proposed solution of Proof of Work blockchain, meaning layers and applications on top of the Bitcoin network. In practice, we see this playing out with initiatives such as the Lightning Network, but also with companies adopting Bitcoin in their services like Square, PayPal, Skrill, Robinhood and many other payment, broker and tech companies.
PoS offers another solution for greater scalability because blocks can be approved in quicker succession, without the need to solve for extremely complex mathematical equations. Randomly choosing the validators based on their stake allows for faster settlement finality with a lower load to achieve this. No reliance on physical machines and external resources to generate consensus in this regard is a positive, as it allows for higher throughput and ultimately more (but not unlimited) scalability.
Decentralization
Many claims are made that PoW blockchains are preferable over PoS, because rewards are not paid out to the coin holders themselves, and thus achieving higher amounts of decentralization in the network and distribution of wealth. While true, the competitive nature of Bitcoin means that miners are very specialized and work at scale. On top of that, often smaller miners pool their computation power to have a higher chance of being selected for producing a block, negating many of the decentralization benefits of Proof of Work.
However, PoS blockchains have similar problems with stake participants handing over their power to a staking provider who does all the validation work for them. Ethereum, the leading PoS protocol, with the most amount of dollar value staked, can be controlled by just three staking providers. Additionally, rewarding participants of the network in the way that PoS does, suggests that the richer get richer over time. It can be argued though, that this is a design choice in how to implement Proof of Stake, and not necessarily a core property. However, early participants in a PoS network are by definition better off than late participants that need to purchase their stake of the early participants to even be allowed to partake in the network.
Security
Proof of Work has some appealing security properties over Proof of Stake. PoW blockchains are physically anchored, where the generation of blocks have a hook into real life resources. Generating blocks requires efficient energy as well as expenditures in specialized hardware mining. This means that the validity of the chain depends partially on external resources. Proof of Stake on the other hand uses an internal resource to secure its chain, namely its own crypto-asset of the network. In this regard, it is easier to create a PoS blockchain, however, in practice there isn't compelling evidence that this difference is material to the security of a blockchain.
The discussion if the security of PoS is comparable to PoW has had considerable debate. The reason for this is that decentralization is similar or roughly indistinguishable between PoW and PoS, and scalability is higher in Proof of Stake systems. From these properties follow, that within the framework of the trilemma that Proof of Work is likely to be more secure than Proof of Stake. However, an apples-to-apples comparison between the two is difficult, because the security of the chain relies on different assumptions to begin with.
In some ways, the discussion gets less important over time, as PoS blockchains are powering hundreds of billions of dollars in value on several PoS blockchains already. Thus, empirical evidence suggests that the security of PoS blockchains is at least ‘good enough’ since no major attack on the consensus of a PoS blockchain has materialized in environments comparable to PoW blockchains with arguably more attack vectors.
Conclusion
The debate about the best blockchain and consensus mechanisms surrounding it will likely continue for the foreseeable future. Even within each consensus mechanism, there are many flavors, with slightly different trade-offs in their designs. All major utility blockchains to date are running a Proof of Stake consensus mechanism, with Ethereum ready to fully move-over to Proof of Stake this September.
Finally, while investing in different blockchains with their properties, an investor needs to keep into account in what part of the technology stack the majority of the value accrual happens since the blockchain ecosystem does not only exists out of Layer-1 blockchains, but also middleware applications and decentralized applications build on top of Layer-1 blockchains.
Fourstack is an investor in all three layers, selecting blockchains, middleware, and decentralized applications on a range of attributes, including the consensus mechanism used.
